Article sorting indication method and device based on bar codes

Abstract

The invention discloses an article sorting indication method and device based on a bar code, and the method comprises the following steps: collecting a panoramic image of a sorting table on which an article to be sorted is placed, and enabling a bar code identifier to be located on the surface, close to an image collection device, of the article to be sorted; segmenting the panoramic image, extracting a bar code identifier in the panoramic image, and calculating a coordinate of a bar code identifier area on the sorting table; analyzing the extracted bar code identifier to generate a corresponding code value, and obtaining a corresponding indication mark according to the code value; synthesizing the bar code identification coordinate and the indication sign through an image synthesis algorithm; and the synthesized image is projected to the to-be-sorted objects on the sorting table. According to the invention, an operator or a robot can quickly identify a sorting target, and the sorting efficiency and accuracy are improved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of article sorting, and particularly relates to an article sorting indication method and device based on barcodes. BACKGROUND

[0002] In the current rapidly developing fields of logistics, express delivery and industrial production, efficient and accurate sorting technology is a key link to improve overall operational efficiency. Traditional sorting methods mainly rely on manual identification, mechanical sorting equipment or simple barcode scanning technology. Barcodes, including one-dimensional barcodes and two-dimensional codes, are widely used as information carriers. One-dimensional barcodes (such as Code 128, EAN-13) encode digital or character information through the width and spacing of black and white stripes, and have limited storage capacity. Two-dimensional codes (such as QR Code, Data Matrix) use matrix two-dimensional encoding and have large storage capacity, allowing them to contain larger amounts of data.

[0003] In the field of express delivery sorting, for example, the sorting flow of express packages in a transfer site is mainly dependent on the following methods: manual PDA terminal scanning, operators need to scan the barcodes on each package one by one, and then manually sort according to system prompts. Due to the single-piece operation, the sorting speed is limited by the efficiency of manual operation, making it difficult to meet the sorting needs of a large number of packages during peak periods. Long hours of high-intensity work can easily lead to operator fatigue, increasing the error rate of mis-sorting and missed scanning, and affecting the overall sorting accuracy. Some large transfer sites use "dynamic weighing + volume measurement + barcode scanning (Dimension Weight Scanning, DWS)" all-in-one machines for pre-sorting, but due to the poor adaptability of the equipment to irregular packages (soft packages, irregular shapes), manual intervention is still required when the barcode is damaged or obscured. Full-automatic sorting lines rely on high-precision barcode recognition and mechanical control, with high construction and maintenance costs, making it difficult for small and medium-sized transfer sites to popularize. In addition, there are strict restrictions on the size, weight and placement of packages, making it impossible to handle non-standardized packages.

[0004] Chinese patent application CN119406761A discloses an artificial-assisted sorting system based on PLC control and a method thereof, which includes a code scanning gun, a PLC controller, a reminding device and an LED display screen. The PLC controller is signal connected with the code scanning gun, the reminding device and the LED display screen. The code scanning gun is used to scan the code of a package and send the obtained package barcode information to the PLC controller. The PLC controller is signal connected with the code scanning gun, and automatically interfaces with a logistics system after receiving the package barcode information to request the bin number of the destination of the goods. The logistics management receives the request and feeds back the bin number of the destination of the package.

[0005] The existing sorting system usually relies on fixed indicator lights or mechanical arms to perform sorting, which cannot adjust the sorting path in real time or provide intuitive operation guidance, affecting the sorting flexibility and accuracy. SUMMARY

[0006] The present application provides a kind of based on bar code article sorting indication method and device, to solve the problems of insufficient sorting flexibility and accuracy of prior art, and operation guidance is not intuitive.

[0007] ‌The bar code described in this embodiment includes one-dimensional code and two-dimensional code.‌Bar code is a kind of black bar and blank with different widths arranged according to certain coding rules to express a group of information. Common bar codes include one-dimensional code (such as EAN-13, Code 128, etc.) and two-dimensional code (such as QR code, Data Matrix, PDF417, etc.).‌

[0008] To solve the above technical problems, the sorting indication method provided by the present application comprises the following steps: Collect the panoramic image of the sorting table on which the to-be-sorted articles are placed, and the bar code mark is located on the surface of the to-be-sorted articles close to the image acquisition device; Segment the panoramic image, extract the bar code mark in the panoramic image, and calculate the coordinates of the bar code mark region on the sorting table; Analyze the extracted bar code mark to generate the corresponding code value, and obtain the corresponding indication mark according to the code value; Synthesize the bar code mark coordinates and the indication mark through an image synthesis algorithm; Project the synthesized image onto the to-be-sorted articles on the sorting table.

[0009] Preferably, the coordinates of the bar code mark region on the sorting table are calculated by the following method: Pretreat the obtained image, including denoising, contrast enhancement, brightness adjustment and histogram equalization; Extract the bar code edge information of the pretreated image using an edge detection algorithm; Identify and segment the bar code region in the image based on the bar code edge information; After segmentation, calculate the center of gravity position of the bar code and the coordinate information of the bar code in the two-dimensional space of the panoramic image using a geometric analysis method, output the corresponding pixel coordinates, and convert the pixel coordinates to physical coordinates corresponding to the actual physical coordinate system of the sorting table.

[0010] Preferably, the extraction method of the bar code edge information is as follows: Calculate the gradient amplitude and direction of the pretreated image using different scales of convolution kernels respectively to generate a gradient amplitude graph; Based on the dynamic allocation of weights according to the energy distribution of gradients at each scale in the gradient magnitude map, a weighted fusion strategy is adopted to fuse gradient magnitudes at different scales to generate a unified edge response map. The edge response map is compared with the pixel neighborhood based on the gradient direction, and an adaptive threshold is set in combination with local statistical parameters to perform edge preservation. A dual-threshold strategy is used to connect and filter edge results. By dynamically adjusting the edge intensity response and tracking effective edges through high and low thresholds, a continuous closed barcode boundary is finally formed.

[0011] Preferably, the method for setting the high and low thresholds is as follows: The mean gradient response is calculated for each of the multiple regions into which the image is divided. with standard deviation According to the formula Set high threshold ,in This is the first adaptive coefficient; According to the high threshold Set low threshold ,satisfy ,in The second adaptive coefficient and .

[0012] Preferably, the method for generating the edge response map is as follows: Calculate the gradient energy value corresponding to each scale, where the gradient energy value is the sum of the squares of the pixel amplitudes in the gradient amplitude map at that scale; The weighting coefficients are calculated based on the normalization of gradient energy values ​​at each scale. The gradient magnitude maps at each scale are weighted and superimposed according to the corresponding weighting coefficients to generate a fused edge response map, which is used for subsequent edge extraction processing.

[0013] Preferably, the size of the convolution kernel includes 3×3, 5×5 and 7×7.

[0014] Preferably, the method for obtaining the indicator is as follows: Based on the corresponding barcode background image, select an easily recognizable color to render the image file extracted from the repository; The size of the rendered image file is adjusted based on the size of the barcode identification area; The position of the rendered image file is adjusted based on the coordinates of the barcode identification area; Save the adjusted image file and generate indicator signs.

[0015] Preferably, the processing flow of the image synthesis algorithm is as follows: The indicator sign is blended with the original image using channel compositing or layer overlay; Mapping the barcode identification area coordinates to indicate the position of the marker; The synthesized image is outputted as a high-resolution file after brightness, contrast and color saturation optimization.

[0016] Preferably, the code value is further subjected to data verification, and the verification method comprises CRC verification, length consistency detection and database comparison verification.

[0017] Another aspect of the present application also provides a barcode-based article sorting indication device, which comprises a sorting table, a support, an image acquisition module and an industrial personal computer, the sorting table is used for placing articles to be sorted and accommodating the industrial personal computer, the support is fixed on the sorting table and used for supporting the image acquisition module, and the image acquisition module is used for acquiring a panoramic image of the sorting table and sending the panoramic image to the industrial personal computer for processing. The device is characterized in that it further comprises a projection module arranged on the support, the projection module projects the synthesized image to a set position of the article to be sorted according to an instruction of the industrial personal computer. The industrial personal computer comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, and when the computer program is executed by the processor, the above-mentioned barcode-based article sorting indication method is realized.

[0018] Compared with the prior art, the present application has the following technical effects: 1. The sorting indication method proposed in the present application extracts barcode information on the article to be sorted dynamically through image acquisition and barcode coordinate identification, calculates the actual physical position of the article to be sorted on the sorting table, realizes accurate positioning of each article, and then synthesizes and projects an indication marker (such as a graphic, color or text prompt) associated with the barcode information directly to the position of the article, so that the operator or robot can identify the sorting target at a glance, thereby improving the sorting efficiency and accuracy.

[0019] 2. The sorting indication method proposed in the present application introduces a multi-scale adaptive gradient edge detection algorithm, combines image preprocessing and double-threshold edge connection technology, and effectively improves the barcode recognition success rate under complex background or low-quality image. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1 is a flowchart of the sorting indication method described in the present application; Figure 2 is an example diagram of the panoramic image described in the embodiment of the present application; Figure 3 is an example diagram of the indication marker described in the embodiment of the present application; Figure 4is the effect drawing after projection of the embodiment of the present application; Figure 5 is a structural schematic diagram of the sorting indication device of the embodiment of the present application; Figure 6 is an exploded view of the sorting indication device of the embodiment of the present application.

[0021] The reference signs: 1, sorting table; 2, support; 3, image acquisition module; 4, industrial host computer; 5, projection module; 11, panel; 12, left baffle; 13, right baffle; 14, rear baffle; 15, side sealing plate; 16, front sealing plate; 17, rear door plate; 18, bottom plate; 19, caster. DETAILED DESCRIPTION

[0022] In order to make the purpose, technical scheme and advantages of the present application more clear, the technical scheme of the present application will be described clearly and completely below by combining with the specific embodiments of the present application and referring to the drawings.

[0023] Embodiment one The present embodiment is a kind of based on bar code's article sorting indication method, as Figure 1 shown, including the following steps one to step five: Step one, collect the full view image of the sorting table on which the to-be-sorted articles are placed, and the bar code mark is located on the surface of the to-be-sorted article close to the image acquisition device.

[0024] In this step, first, there is a standard size sorting table, and the table surface can be made of matte material to reduce glare interference. A industrial vision camera is vertically installed above the sorting table, and the camera is fixed through a cantilever structure, at a proper height from the sorting table surface to facilitate focusing and ensure that the full view image of the entire sorting area can be shot.

[0025] Before sorting operation, the staff randomly places multiple to-be-sorted articles on the sorting table surface, and a bar code mark or a two-dimensional code mark is pasted on the outer packaging of each article. In order to ensure the subsequent image recognition effect, the operator needs to manually rotate the article so that each bar code or two-dimensional code faces upward, i.e. the surface close to the image acquisition device. Or set a computer vision module and a corresponding mechanical turnover module on the conveying device in front of the sorting table 1, and cooperate to turn over to realize the bar code face upward. The bar code surface should be in an unobstructed state, and wrinkles, obstructions and reflections should be avoided as much as possible.

[0026] In addition, the collection environment is equipped with a light supplementing device such as a ring-shaped soft light LED light strip to avoid bar code area recognition failure due to shadows or overexposure. Before collection, the camera sends a trigger signal through a software interface and collects images in a stable light and relatively static state to generate a high-resolution sorting table overhead view, i.e. the so-called "full view image", as Figure 2An example of the panoramic image collected in this embodiment is shown.

[0027] In this embodiment, the camera can access the control system through an external industrial host and can automatically collect images according to a preset time interval or a trigger signal. The resulting images are stored in a lossless format such as PNG or TIFF and serve as input data for subsequent image processing modules.

[0028] By using high-position visual imaging equipment and combining with barcode orientation control operations, a complete sorting table panoramic image containing multiple items to be sorted is collected, meeting the requirements of image clarity, resolution and barcode orientation accuracy, and providing high-quality image data basis for subsequent segmentation, coordinate recognition and projection indication.

[0029] Step two, segmenting the panoramic image, extracting the barcode identifier in the panoramic image, and calculating the coordinates of the barcode identifier region on the sorting table.

[0030] In this step, the coordinates of the barcode identifier region on the sorting table are calculated by the following method: The acquired image is preprocessed, including denoising, contrast enhancement, brightness adjustment and histogram equalization. The denoising process uses a Gaussian filter (such as σ=0.5) to smooth the original image and remove noise points. The contrast enhancement process uses a linear stretching method to enhance the contrast of the image and highlight the boundaries between the objects and the background. The brightness adjustment process globally or locally corrects the brightness according to the image gray mean value. The histogram equalization process equalizes the image gray distribution and improves the gray variation significance of the barcode region.

[0031] An edge detection algorithm is used to extract the barcode edge information of the preprocessed image. The edge detection algorithm is used to extract the structural features of the barcode boundary. The method used is a multi-scale adaptive gradient edge detection algorithm Based on the barcode edge information, the barcode region in the image is recognized and segmented; After segmentation, the barycenter position of the barcode and the coordinate information of the barcode in the two-dimensional space of the panoramic image are calculated using a geometric analysis method, the corresponding pixel coordinates are output, and the pixel coordinates are converted into physical coordinates corresponding to the actual physical coordinate system of the sorting table.

[0032] The method for extracting the barcode edge information is as follows: Different scales of convolution kernels are used to calculate the gradient amplitude and direction of the preprocessed image to generate a gradient amplitude map. In this embodiment, the scales of the convolution kernels include 3x3, 5x5 and 7x7. Three scales of Sobel convolution kernels, i.e., 3x3, 5x5 and 7x7, are used to calculate the gradients in the X and Y directions of the image, and the gradient amplitude map and direction map under each scale are calculated accordingly.

[0033] According to the energy distribution of the gradient of each scale in the gradient amplitude map, a weighted fusion strategy is adopted to fuse the gradient amplitudes of different scales to generate a unified edge response map; the generation method of the edge response map is specifically as follows: the gradient energy value corresponding to each scale is calculated respectively, and the gradient energy value is the sum of the squares of the amplitudes of each pixel in the gradient amplitude map of the scale; the weighted coefficients are calculated based on the normalized gradient energy values of each scale; the gradient amplitude maps of each scale are weighted and superimposed according to the corresponding weighted coefficients to generate a fused edge response map, which is used for subsequent edge extraction processing.

[0034] According to the pixel neighborhood comparison of the gradient direction of the edge response map, and combining with the setting of the adaptive threshold based on the local statistical parameters, the edge reservation is performed; specifically, for each pixel in the edge response map, it is divided into four main directions (0°, 45°, 90°, 135°) according to its gradient direction, and comparison is made in the neighborhood, and only the pixel value with the same direction and local maximum is reserved as the candidate edge point.

[0035] A double-threshold strategy is adopted to connect and screen the edge results, the edge intensity response is dynamically adjusted through high and low thresholds, and the effective edge is tracked, and finally the continuous and closed barcode boundary is formed. In this embodiment, a Canny type edge tracking mechanism is adopted, starting from the edge point with high threshold response, connecting all the adjacent effective edge points with low threshold response, and finally forming a complete and closed barcode boundary.

[0036] The setting method of the high and low thresholds is as follows: In the multiple regions divided in the image, the mean value of the gradient response in each region is calculated and the standard deviation , according to the formula the high threshold is set , wherein is the first adaptive coefficient; According to the high threshold the low threshold is set , which satisfies , wherein is the second adaptive coefficient and .

[0037] After edge extraction, barcode region recognition and contour segmentation are performed based on the closed boundary to extract the rectangular region containing the barcode in the image. Then the following coordinate calculation process is performed: barycenter extraction, the barycenter pixel coordinates (x, y) of each barcode region are calculated through boundary geometric shape analysis; coordinate conversion, the pixel coordinates are mapped to the physical coordinate system of the sorting table, and the transformation from pixel to physical coordinates is completed by using the camera calibration parameters (such as intrinsic matrix, distortion coefficient and external parameter pose), and the coordinates (X_real, Y_real) of the barcode center in the actual physical space are output.

[0038] Step three, parse the extracted barcode identifier to generate the corresponding code value, and obtain the corresponding indication mark according to the code value.

[0039] The barcode image area extracted in step two is parsed, and a general barcode recognition algorithm (such as ZBar, OpenCV or self-developed recognition model) is used to decode the one-dimensional code or two-dimensional code content in the image, and the code value corresponding to the barcode is extracted.

[0040] In the parsing process, this embodiment also introduces a data verification mechanism to improve the reliability of the recognition result: if the barcode is of a custom format, perform cyclic redundancy check (CRC-16 or CRC-32) after reading the code value to determine whether there is a transmission or image recognition error in the parsed content; according to the preset rule, whether the character length of the verification code value conforms to the length of the corresponding business field (such as customer code for 10 bits, order number for 16 bits, etc.); match and verify the parsed code value with the valid code value in the background database to confirm its legality and current state.

[0041] If any of the above verifications fails, the system will automatically execute a retry mechanism, including re-invoking the barcode recognition function, re-acquiring the image of the area and retrying; if it fails multiple times, the system will generate an error prompt flag to prompt the operator for manual intervention.

[0042] The method for obtaining the indication mark is: According to the corresponding barcode identifier background image, select an easily recognizable color to render the image file extracted from the storage library; adjust the size of the rendered image file based on the size of the barcode identifier area; adjust the position of the rendered image file based on the coordinates of the barcode identifier area; save the adjusted image file to generate the indication mark. Specifically, when the code value is verified to be valid, the system generates corresponding image guidance information (i.e. "indication mark") according to the code value, and the specific generation process is as follows: Call the corresponding image file from the storage library: according to the classification rules (such as the type of goods, priority, sorting path, etc.) corresponding to the code value, call the indication image (such as left arrow, right arrow, red box, text identification, etc.) corresponding to the type of goods in the local or remote indication icon database.

[0043] Color rendering optimization: combined with the background image of the barcode area (analyzed by the image pixels near the barcode area), select the color scheme with the strongest contrast and easiest to distinguish to render the selected image file (for example: use dark arrow marks on light background, automatically switch to black or dark red tone).

[0044] Size adjustment: according to the physical size of the barcode identification area, the rendered instruction icon is scaled to coordinate with the actual size of the barcode when projected onto the sorting table, neither blocking the content nor ensuring easy identification. The scaling rule can be automatically set according to the preset proportion (for example: the instruction sign width is 1.5 times the barcode width).

[0045] Position adjustment and fusion: align the rendered image according to the physical coordinates of the barcode area, align the center point with the barycentric coordinates of the barcode, or slightly offset upwards so that the barcode content does not overlap with the instruction sign (the offset is dynamically calculated by the system), and realize the visual association of the guide image and the barcode.

[0046] Save and output: cache the adjusted image file to the image synthesis module and mark it as the "instruction sign" resource corresponding to the barcode for subsequent image synthesis and projection module. The final instruction sign is as shown in Figure 3 Figure 3 As two instruction signs, they correspond to different background colors respectively.

[0047] Through multiple verification of code value, the accuracy and stability of analysis are improved, and based on background color, coordinate and size, adaptive rendering and image adjustment are performed, and the generated instruction sign has good visibility and human-computer interaction experience, and is suitable for sorting scenes under different lighting and background conditions.

[0048] Step four, through image synthesis algorithm, the barcode identification coordinates and the instruction sign are synthesized. Based on the instruction sign image generated in step three and the barcode identification area coordinates extracted in step two, the two are fused through image synthesis algorithm to generate a synthesized image for subsequent projection. The processing flow of the image synthesis algorithm is as follows: Fuse the instruction sign and the original image through channel synthesis or layer superposition: This embodiment preferably uses an image synthesis method based on layer superposition mechanism to realize the fusion of barcode identification and instruction sign image. This method is completed through the following processing flow: construct the original image and the instruction sign image as two independent layers; set the instruction sign image as a transparent background (alpha channel control) to support not blocking the barcode information; use image superposition function (such as cv2.addWeighted() in OpenCV or paste() function in PIL library) to perform pixel-level synthesis of the two layers; support on-demand selection of different channel (RGB or grayscale) synthesis mode, compatible with color and black and white projection scenes.

[0049] Map the instruction sign position based on the barcode identification area coordinates: ​To ensure that the indicator in the synthetic image is accurately displayed in the target area, the following positioning strategy is adopted in this embodiment: reading the barcode barycenter pixel coordinates (x, y) obtained in step two; determining the starting position (top-left corner coordinates) of the inserted layer of the indicator as (x-w / 2, y-h / 2) according to the adjusted size of the indicator in step three, wherein w and h are the width and height of the indicator image respectively; if the barcode is located in the edge area of the image, the system automatically adjusts the insertion coordinates to prevent the indicator image from exceeding the boundary; supporting the rotation matching function, adjusting the rotation angle of the indicator according to the direction of the barcode to make it consistent or perpendicular to the direction of the barcode, thereby improving the guiding and identifying efficiency.

[0050] After brightness, contrast and color saturation optimization of the synthetic image, a high-resolution file is output: To ensure the display effect of the final synthetic image in the projection device, the following image enhancement operations are further performed after image synthesis in this embodiment: brightness enhancement, linear stretching or Gamma correction (default γ=1.2-1.6) according to the overall brightness distribution of the image; contrast adjustment, histogram equalization is used to enhance the contrast of the edge area, so that the barcode and the edge of the indicator are more clear; color saturation enhancement, S channel (saturation) is enhanced in the RGB space to improve the color differentiation, which is particularly suitable for scenes where multiple indicators are projected simultaneously.

[0051] In some other embodiments of the present application, the image enhancement operation further includes resolution enhancement, specifically using bilinear interpolation or super-resolution model to enlarge the synthetic image (the default output resolution is 1080p or 4K) to meet the clarity requirements of different projection devices.

[0052] After synthesis is completed, the image file is cached to a temporary directory with a unified naming rule for subsequent step five projection module to read and output.

[0053] In step five, the synthesized image is projected onto the to-be-sorted items on the sorting table. Figure 4 As shown in the figure, this embodiment uses a high-speed short-focus LED projection device to refresh the image at a fixed frequency (such as 15-30 frames per second) through the control system to realize real-time dynamic projection.

[0054] In some other embodiments of the present application, to realize accurate projection, the system first establishes the spatial coordinate mapping relationship between the projection device and the sorting table, which specifically includes: using a calibration board to perform spatial geometric calibration to obtain the transformation matrix (such as a homogeneous matrix) between the projector and the image acquisition device; constructing an affine transformation model or a perspective transformation model from the image pixel coordinate system to the physical coordinate system of the sorting table to ensure that the projection position of the image content in the physical space is aligned with the actual position of the barcode; using single-point or multi-point mapping to optimize the coordinate conversion accuracy in real time to adapt to the slight changes in the posture of the projector or the height of the workbench.

[0055] Before the image is output, the system processes the synthesized image according to the calibration data, such as performing affine or perspective transformation on the image according to the projection transformation matrix to eliminate image distortion caused by the projection angle; and occlusion region processing: when it is detected that the barcode region is occluded (such as a cover paper or a hand), the system does not output the image temporarily to avoid misguiding.

[0056] Embodiment Two The present embodiment is a barcode-based article sorting indication device, as shown in Figure 5 , 6 The device comprises a sorting table 1, a support 2, an image acquisition module 3, and an industrial personal computer 4. The sorting table 1 is used to place articles to be sorted and accommodate the industrial personal computer 4. The support 2 is fixed on the sorting table 1 and is used to support the image acquisition module 3. The image acquisition module 3 is used to acquire a panoramic image of the sorting table 1 and send it to the industrial personal computer 4 for processing. The device is characterized in that it further comprises a projection module 5 arranged on the support. The projection module 5 projects the synthesized image to a set position of the article to be sorted according to the instruction of the industrial personal computer 4. The industrial personal computer comprises a memory, a processor, and a computer program stored on the memory and executable on the processor. When the computer program is executed by the processor, the barcode-based article sorting indication method as described in Embodiment One is realized.

[0057] Specifically, the sorting table 1 comprises a panel 11, a baffle, and a frame. In the present embodiment, the frame is a cuboid frame used to fix the panel 11, the baffle, and an enclosing plate and form a clamping key to accommodate the industrial personal computer 4. The panel 11 is square in structure, and the size of the panel 11 is 1 meter*1 meter. The baffle is divided into a left baffle 12, a right baffle 13, and a rear baffle 14. The left baffle 12 is connected to the left side of the panel 11, the right baffle 13 is connected to the right side of the panel 11, and the rear baffle 14 is connected to the rear side of the panel 11. The baffle is used to prevent the article to be sorted from falling out of the panel 1. The frame is made of aluminum alloy, and the height is 0.75 meters. The sorting table 1 is used to temporarily store the article to be sorted. An operator can stand in front of the sorting table to check the express package to ensure that the barcode information on the package is upward and not occluded.

[0058] In some other embodiments of the present application, the sorting table 1 is further provided with a side enclosing plate 15, a front enclosing plate 16, a rear door plate 17, and a bottom plate 18 surrounding the frame. The side enclosing plate 15 comprises a left side enclosing plate and a right side enclosing plate. The front enclosing plate 16 is provided with an opening corresponding to the position of the operation button or IO interface of the industrial personal computer 4.

[0059] In some other embodiments of the present application, casters 19 are also arranged below the frame.

[0060] The image acquisition module 3 and the projection module 5 are arranged on the support 2, specifically, are fixed above the panel 11 by the support 2, and when used, the to-be-sorted items are placed on the panel 11 with the barcodes facing upward. The to-be-sorted items are imaged by the image acquisition module 3 above and then transmitted to the industrial personal computer 4. In addition, the projection module 5 above the panel 11 projects the synthesized image onto the to-be-sorted items on the panel 11, forming an intuitive sorting instruction.

[0061] The operation mechanism of the device is as follows: Sorting and arranging, the operator stands in front of the sorting table 1, arranges the packages on the sorting table 1, and ensures that the packages are not overlapped and the barcodes of the packages face upward; or a computer vision module and a corresponding mechanical turnover module are arranged on the conveying device in front of the sorting table 1, which are matched to realize the turnover of the barcodes facing upward.

[0062] Start the image acquisition module 3 block, ensure that all the packages on the sorting table 1 are on the photo, perform image acquisition and identification on the packages, and transmit the acquired image to the industrial personal computer 4 for visual algorithm operation.

[0063] According to the visual recognition result of the package by the software of the industrial personal computer 4, the length and width two-dimensional dimensions and the two-dimensional space coordinates in the visual coordinate system of the package on the sorting table are obtained; the original image is processed by a visual algorithm to obtain the code value of the barcode or two-dimensional code and the two-dimensional coordinates of each barcode or two-dimensional code.

[0064] The decoding algorithm analyzes the barcode or two-dimensional code and outputs the corresponding code value.

[0065] According to the code value, an indication mark is generated, which is bright in color and located at the upper right of the barcode or two-dimensional code, so as to facilitate personnel to quickly identify.

[0066] The indication mark is synthesized into the corresponding barcode coordinate of the original image.

[0067] The above only describes the preferred embodiments of the present application, and it should be noted that for ordinary skilled persons in the art, without departing from the inventive concept, a number of modifications and improvements can be made, which are all within the protection scope of the present application.

Claims

1. A barcode-based item sorting instruction method, characterized in that, Includes the following steps: Capture a panoramic image of the sorting table where items to be sorted are placed, with barcode labels located on the surface of the items to be sorted near the image acquisition device; The panoramic image is segmented, the barcode identifier in the panoramic image is extracted, and the coordinates of the barcode identifier area on the sorting table are calculated; The extracted barcode identifier is parsed to generate the corresponding code value, and the corresponding indicator is obtained based on the code value; The barcode coordinates and indicator marks are combined using an image synthesis algorithm. The composite image is projected onto the items to be sorted on the sorting table.

2. The method according to claim 1, characterized in that, The specific method for calculating the coordinates of the barcode identification area on the sorting table is as follows: The acquired images are preprocessed, including noise reduction, contrast enhancement, brightness adjustment, and histogram equalization. An edge detection algorithm is used to extract barcode edge information from the preprocessed image; Barcode regions in images are identified and segmented based on barcode edge information; After segmentation, the barcode's centroid position and coordinate information in the two-dimensional space of the panoramic image are calculated using geometric analysis methods. The corresponding pixel coordinates are then output and converted into physical coordinates corresponding to the actual physical coordinate system of the sorting station.

3. The method according to claim 2, characterized in that, The method for extracting the barcode edge information is as follows: The gradient magnitude and direction of the preprocessed image are calculated using convolution kernels of different scales, and a gradient magnitude map is generated. Based on the dynamic allocation of weights according to the energy distribution of gradients at each scale in the gradient magnitude map, a weighted fusion strategy is adopted to fuse gradient magnitudes at different scales to generate a unified edge response map. The edge response map is compared with the pixel neighborhood based on the gradient direction, and an adaptive threshold is set in combination with local statistical parameters to perform edge preservation. A dual-threshold strategy is used to connect and filter edge results. By dynamically adjusting the edge intensity response and tracking effective edges through high and low thresholds, a continuous closed barcode boundary is finally formed.

4. The method according to claim 3, characterized in that, The method for setting the high and low thresholds is as follows: The mean gradient response is calculated for each of the multiple regions into which the image is divided. with standard deviation According to the formula Set high threshold ,in This is the first adaptive coefficient; According to the high threshold Set low threshold ,satisfy ,in The second adaptive coefficient and .

5. The method according to claim 3, characterized in that, The method for generating the edge response map is as follows: Calculate the gradient energy value corresponding to each scale, where the gradient energy value is the sum of the squares of the pixel amplitudes in the gradient amplitude map at that scale; The weighting coefficients are calculated based on the normalization of gradient energy values ​​at each scale. The gradient magnitude maps at each scale are weighted and superimposed according to the corresponding weighting coefficients to generate a fused edge response map, which is used for subsequent edge extraction processing.

6. The method according to claim 3, characterized in that, The convolution kernels include sizes of 3×3, 5×5, and 7×7.

7. The method according to claim 1, characterized in that, The method for obtaining the indicator is as follows: Based on the corresponding barcode background image, select an easily recognizable color to render the image file extracted from the repository; The size of the rendered image file is adjusted based on the size of the barcode identification area; The position of the rendered image file is adjusted based on the coordinates of the barcode identification area; Save the adjusted image file and generate indicator signs.

8. The method according to claim 1, characterized in that, The processing flow of the image synthesis algorithm is as follows: The indicator sign is blended with the original image using channel compositing or layer overlay; The location of the marker is indicated by mapping the coordinates of the barcode identification area. After optimizing the brightness, contrast, and color saturation of the synthesized image, a high-resolution file is output.

9. The method according to claim 1, characterized in that, The code value is also subject to data verification, including CRC check, length consistency detection and database comparison verification. If the verification fails, a retry mechanism is executed or an error message is generated.

10. A barcode-based item sorting indicator device, comprising a sorting table, a support, an image acquisition module, and an industrial control host computer, wherein the sorting table is used to place items to be sorted and to house the industrial control host computer, the support is fixed on the sorting table to support the image acquisition module, and the image acquisition module is used to acquire panoramic images of the sorting table and send them to the industrial control host computer for processing. Its features are, It also includes a projection module mounted on a bracket, which projects the composite image onto a set position of the items to be sorted according to the instructions of the industrial control host computer. The industrial control host computer includes: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it implements the barcode-based item sorting instruction method as described in any one of claims 1-9.

Images